[BACK]Return to mandoc.c CVS log [TXT][DIR] Up to [cvsweb.bsd.lv] / mandoc

Annotation of mandoc/mandoc.c, Revision 1.82

1.82    ! schwarze    1: /*     $Id: mandoc.c,v 1.81 2014/07/01 22:37:15 schwarze Exp $ */
1.1       kristaps    2: /*
1.59      schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.78      schwarze    4:  * Copyright (c) 2011, 2012, 2013, 2014 Ingo Schwarze <schwarze@openbsd.org>
1.1       kristaps    5:  *
                      6:  * Permission to use, copy, modify, and distribute this software for any
                      7:  * purpose with or without fee is hereby granted, provided that the above
                      8:  * copyright notice and this permission notice appear in all copies.
                      9:  *
1.36      schwarze   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.1       kristaps   11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.36      schwarze   12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.1       kristaps   13:  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
                     14:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
                     15:  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
                     16:  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
                     17:  */
1.9       kristaps   18: #ifdef HAVE_CONFIG_H
                     19: #include "config.h"
1.7       kristaps   20: #endif
                     21:
1.2       kristaps   22: #include <sys/types.h>
                     23:
1.1       kristaps   24: #include <assert.h>
                     25: #include <ctype.h>
1.50      kristaps   26: #include <errno.h>
                     27: #include <limits.h>
1.1       kristaps   28: #include <stdlib.h>
1.4       kristaps   29: #include <stdio.h>
                     30: #include <string.h>
1.7       kristaps   31: #include <time.h>
1.1       kristaps   32:
1.18      kristaps   33: #include "mandoc.h"
1.76      schwarze   34: #include "mandoc_aux.h"
1.1       kristaps   35: #include "libmandoc.h"
                     36:
1.37      schwarze   37: #define DATESIZE 32
                     38:
1.18      kristaps   39: static int      a2time(time_t *, const char *, const char *);
1.37      schwarze   40: static char    *time2a(time_t);
1.7       kristaps   41:
1.45      kristaps   42:
                     43: enum mandoc_esc
1.74      schwarze   44: mandoc_escape(const char **end, const char **start, int *sz)
1.1       kristaps   45: {
1.65      schwarze   46:        const char      *local_start;
                     47:        int              local_sz;
                     48:        char             term;
1.79      schwarze   49:        enum mandoc_esc  gly;
1.45      kristaps   50:
1.65      schwarze   51:        /*
                     52:         * When the caller doesn't provide return storage,
                     53:         * use local storage.
                     54:         */
                     55:
                     56:        if (NULL == start)
                     57:                start = &local_start;
                     58:        if (NULL == sz)
                     59:                sz = &local_sz;
                     60:
                     61:        /*
                     62:         * Beyond the backslash, at least one input character
                     63:         * is part of the escape sequence.  With one exception
                     64:         * (see below), that character won't be returned.
                     65:         */
                     66:
1.45      kristaps   67:        gly = ESCAPE_ERROR;
1.65      schwarze   68:        *start = ++*end;
                     69:        *sz = 0;
1.64      schwarze   70:        term = '\0';
1.18      kristaps   71:
1.65      schwarze   72:        switch ((*start)[-1]) {
1.45      kristaps   73:        /*
                     74:         * First the glyphs.  There are several different forms of
                     75:         * these, but each eventually returns a substring of the glyph
                     76:         * name.
                     77:         */
1.79      schwarze   78:        case '(':
1.45      kristaps   79:                gly = ESCAPE_SPECIAL;
1.65      schwarze   80:                *sz = 2;
1.45      kristaps   81:                break;
1.79      schwarze   82:        case '[':
1.45      kristaps   83:                gly = ESCAPE_SPECIAL;
1.52      kristaps   84:                /*
                     85:                 * Unicode escapes are defined in groff as \[uXXXX] to
                     86:                 * \[u10FFFF], where the contained value must be a valid
                     87:                 * Unicode codepoint.  Here, however, only check whether
                     88:                 * it's not a zero-width escape.
                     89:                 */
1.65      schwarze   90:                if ('u' == (*start)[0] && ']' != (*start)[1])
1.52      kristaps   91:                        gly = ESCAPE_UNICODE;
1.45      kristaps   92:                term = ']';
                     93:                break;
1.79      schwarze   94:        case 'C':
1.65      schwarze   95:                if ('\'' != **start)
1.45      kristaps   96:                        return(ESCAPE_ERROR);
1.65      schwarze   97:                *start = ++*end;
1.70      schwarze   98:                if ('u' == (*start)[0] && '\'' != (*start)[1])
                     99:                        gly = ESCAPE_UNICODE;
                    100:                else
                    101:                        gly = ESCAPE_SPECIAL;
1.45      kristaps  102:                term = '\'';
                    103:                break;
1.72      schwarze  104:
                    105:        /*
                    106:         * Escapes taking no arguments at all.
                    107:         */
1.79      schwarze  108:        case 'd':
1.72      schwarze  109:                /* FALLTHROUGH */
1.79      schwarze  110:        case 'u':
1.72      schwarze  111:                return(ESCAPE_IGNORE);
1.63      schwarze  112:
                    113:        /*
                    114:         * The \z escape is supposed to output the following
1.79      schwarze  115:         * character without advancing the cursor position.
1.63      schwarze  116:         * Since we are mostly dealing with terminal mode,
                    117:         * let us just skip the next character.
                    118:         */
1.79      schwarze  119:        case 'z':
1.63      schwarze  120:                return(ESCAPE_SKIPCHAR);
1.1       kristaps  121:
1.45      kristaps  122:        /*
                    123:         * Handle all triggers matching \X(xy, \Xx, and \X[xxxx], where
                    124:         * 'X' is the trigger.  These have opaque sub-strings.
                    125:         */
1.79      schwarze  126:        case 'F':
1.45      kristaps  127:                /* FALLTHROUGH */
1.79      schwarze  128:        case 'g':
1.24      kristaps  129:                /* FALLTHROUGH */
1.79      schwarze  130:        case 'k':
1.24      kristaps  131:                /* FALLTHROUGH */
1.79      schwarze  132:        case 'M':
1.24      kristaps  133:                /* FALLTHROUGH */
1.79      schwarze  134:        case 'm':
1.24      kristaps  135:                /* FALLTHROUGH */
1.79      schwarze  136:        case 'n':
1.24      kristaps  137:                /* FALLTHROUGH */
1.79      schwarze  138:        case 'V':
1.24      kristaps  139:                /* FALLTHROUGH */
1.79      schwarze  140:        case 'Y':
1.60      schwarze  141:                gly = ESCAPE_IGNORE;
1.24      kristaps  142:                /* FALLTHROUGH */
1.79      schwarze  143:        case 'f':
1.45      kristaps  144:                if (ESCAPE_ERROR == gly)
                    145:                        gly = ESCAPE_FONT;
1.65      schwarze  146:                switch (**start) {
1.79      schwarze  147:                case '(':
1.65      schwarze  148:                        *start = ++*end;
                    149:                        *sz = 2;
1.45      kristaps  150:                        break;
1.79      schwarze  151:                case '[':
1.65      schwarze  152:                        *start = ++*end;
1.45      kristaps  153:                        term = ']';
                    154:                        break;
                    155:                default:
1.65      schwarze  156:                        *sz = 1;
1.45      kristaps  157:                        break;
                    158:                }
                    159:                break;
                    160:
                    161:        /*
                    162:         * These escapes are of the form \X'Y', where 'X' is the trigger
                    163:         * and 'Y' is any string.  These have opaque sub-strings.
1.78      schwarze  164:         * The \B and \w escapes are handled in roff.c, roff_res().
1.45      kristaps  165:         */
1.79      schwarze  166:        case 'A':
1.24      kristaps  167:                /* FALLTHROUGH */
1.79      schwarze  168:        case 'b':
1.24      kristaps  169:                /* FALLTHROUGH */
1.79      schwarze  170:        case 'D':
1.24      kristaps  171:                /* FALLTHROUGH */
1.79      schwarze  172:        case 'o':
1.24      kristaps  173:                /* FALLTHROUGH */
1.79      schwarze  174:        case 'R':
1.73      schwarze  175:                /* FALLTHROUGH */
1.79      schwarze  176:        case 'X':
1.24      kristaps  177:                /* FALLTHROUGH */
1.79      schwarze  178:        case 'Z':
1.77      schwarze  179:                if ('\0' == **start)
1.45      kristaps  180:                        return(ESCAPE_ERROR);
                    181:                gly = ESCAPE_IGNORE;
1.77      schwarze  182:                term = **start;
1.65      schwarze  183:                *start = ++*end;
1.24      kristaps  184:                break;
1.45      kristaps  185:
                    186:        /*
                    187:         * These escapes are of the form \X'N', where 'X' is the trigger
                    188:         * and 'N' resolves to a numerical expression.
                    189:         */
1.79      schwarze  190:        case 'h':
1.28      kristaps  191:                /* FALLTHROUGH */
1.79      schwarze  192:        case 'H':
1.45      kristaps  193:                /* FALLTHROUGH */
1.79      schwarze  194:        case 'L':
1.45      kristaps  195:                /* FALLTHROUGH */
1.79      schwarze  196:        case 'l':
1.45      kristaps  197:                /* FALLTHROUGH */
1.79      schwarze  198:        case 'S':
1.45      kristaps  199:                /* FALLTHROUGH */
1.79      schwarze  200:        case 'v':
1.28      kristaps  201:                /* FALLTHROUGH */
1.79      schwarze  202:        case 'x':
1.82    ! schwarze  203:                if (strchr(" %&()*+-./0123456789:<=>", **start)) {
        !           204:                        ++*end;
1.65      schwarze  205:                        return(ESCAPE_ERROR);
1.82    ! schwarze  206:                }
1.73      schwarze  207:                gly = ESCAPE_IGNORE;
1.77      schwarze  208:                term = **start;
1.65      schwarze  209:                *start = ++*end;
1.45      kristaps  210:                break;
1.60      schwarze  211:
                    212:        /*
                    213:         * Special handling for the numbered character escape.
                    214:         * XXX Do any other escapes need similar handling?
                    215:         */
1.79      schwarze  216:        case 'N':
1.65      schwarze  217:                if ('\0' == **start)
1.60      schwarze  218:                        return(ESCAPE_ERROR);
1.65      schwarze  219:                (*end)++;
                    220:                if (isdigit((unsigned char)**start)) {
                    221:                        *sz = 1;
1.60      schwarze  222:                        return(ESCAPE_IGNORE);
1.65      schwarze  223:                }
                    224:                (*start)++;
1.60      schwarze  225:                while (isdigit((unsigned char)**end))
                    226:                        (*end)++;
1.65      schwarze  227:                *sz = *end - *start;
1.60      schwarze  228:                if ('\0' != **end)
                    229:                        (*end)++;
                    230:                return(ESCAPE_NUMBERED);
1.45      kristaps  231:
1.79      schwarze  232:        /*
1.45      kristaps  233:         * Sizes get a special category of their own.
                    234:         */
1.79      schwarze  235:        case 's':
1.45      kristaps  236:                gly = ESCAPE_IGNORE;
1.28      kristaps  237:
1.45      kristaps  238:                /* See +/- counts as a sign. */
1.65      schwarze  239:                if ('+' == **end || '-' == **end || ASCII_HYPH == **end)
                    240:                        (*end)++;
1.8       kristaps  241:
1.65      schwarze  242:                switch (**end) {
1.79      schwarze  243:                case '(':
1.65      schwarze  244:                        *start = ++*end;
                    245:                        *sz = 2;
1.22      kristaps  246:                        break;
1.79      schwarze  247:                case '[':
1.65      schwarze  248:                        *start = ++*end;
1.64      schwarze  249:                        term = ']';
1.22      kristaps  250:                        break;
1.79      schwarze  251:                case '\'':
1.65      schwarze  252:                        *start = ++*end;
1.64      schwarze  253:                        term = '\'';
1.22      kristaps  254:                        break;
                    255:                default:
1.65      schwarze  256:                        *sz = 1;
1.22      kristaps  257:                        break;
1.8       kristaps  258:                }
                    259:
1.45      kristaps  260:                break;
1.33      kristaps  261:
1.45      kristaps  262:        /*
                    263:         * Anything else is assumed to be a glyph.
1.65      schwarze  264:         * In this case, pass back the character after the backslash.
1.45      kristaps  265:         */
                    266:        default:
                    267:                gly = ESCAPE_SPECIAL;
1.65      schwarze  268:                *start = --*end;
                    269:                *sz = 1;
1.22      kristaps  270:                break;
1.45      kristaps  271:        }
                    272:
                    273:        assert(ESCAPE_ERROR != gly);
                    274:
                    275:        /*
1.64      schwarze  276:         * Read up to the terminating character,
                    277:         * paying attention to nested escapes.
1.45      kristaps  278:         */
                    279:
                    280:        if ('\0' != term) {
1.64      schwarze  281:                while (**end != term) {
                    282:                        switch (**end) {
1.79      schwarze  283:                        case '\0':
1.64      schwarze  284:                                return(ESCAPE_ERROR);
1.79      schwarze  285:                        case '\\':
1.64      schwarze  286:                                (*end)++;
                    287:                                if (ESCAPE_ERROR ==
                    288:                                    mandoc_escape(end, NULL, NULL))
                    289:                                        return(ESCAPE_ERROR);
                    290:                                break;
                    291:                        default:
                    292:                                (*end)++;
                    293:                                break;
                    294:                        }
                    295:                }
1.65      schwarze  296:                *sz = (*end)++ - *start;
1.64      schwarze  297:        } else {
1.65      schwarze  298:                assert(*sz > 0);
                    299:                if ((size_t)*sz > strlen(*start))
1.45      kristaps  300:                        return(ESCAPE_ERROR);
1.65      schwarze  301:                *end += *sz;
1.45      kristaps  302:        }
                    303:
                    304:        /* Run post-processors. */
                    305:
                    306:        switch (gly) {
1.79      schwarze  307:        case ESCAPE_FONT:
1.68      schwarze  308:                if (2 == *sz) {
                    309:                        if ('C' == **start) {
                    310:                                /*
                    311:                                 * Treat constant-width font modes
                    312:                                 * just like regular font modes.
                    313:                                 */
                    314:                                (*start)++;
                    315:                                (*sz)--;
                    316:                        } else {
                    317:                                if ('B' == (*start)[0] && 'I' == (*start)[1])
                    318:                                        gly = ESCAPE_FONTBI;
                    319:                                break;
                    320:                        }
1.65      schwarze  321:                } else if (1 != *sz)
1.45      kristaps  322:                        break;
1.61      kristaps  323:
1.65      schwarze  324:                switch (**start) {
1.79      schwarze  325:                case '3':
1.45      kristaps  326:                        /* FALLTHROUGH */
1.79      schwarze  327:                case 'B':
1.45      kristaps  328:                        gly = ESCAPE_FONTBOLD;
                    329:                        break;
1.79      schwarze  330:                case '2':
1.45      kristaps  331:                        /* FALLTHROUGH */
1.79      schwarze  332:                case 'I':
1.45      kristaps  333:                        gly = ESCAPE_FONTITALIC;
1.22      kristaps  334:                        break;
1.79      schwarze  335:                case 'P':
1.45      kristaps  336:                        gly = ESCAPE_FONTPREV;
1.22      kristaps  337:                        break;
1.79      schwarze  338:                case '1':
1.45      kristaps  339:                        /* FALLTHROUGH */
1.79      schwarze  340:                case 'R':
1.45      kristaps  341:                        gly = ESCAPE_FONTROMAN;
1.1       kristaps  342:                        break;
                    343:                }
1.46      kristaps  344:                break;
1.79      schwarze  345:        case ESCAPE_SPECIAL:
1.65      schwarze  346:                if (1 == *sz && 'c' == **start)
1.45      kristaps  347:                        gly = ESCAPE_NOSPACE;
1.22      kristaps  348:                break;
1.1       kristaps  349:        default:
1.22      kristaps  350:                break;
1.1       kristaps  351:        }
                    352:
1.45      kristaps  353:        return(gly);
1.36      schwarze  354: }
                    355:
                    356: /*
                    357:  * Parse a quoted or unquoted roff-style request or macro argument.
                    358:  * Return a pointer to the parsed argument, which is either the original
                    359:  * pointer or advanced by one byte in case the argument is quoted.
1.71      schwarze  360:  * NUL-terminate the argument in place.
1.36      schwarze  361:  * Collapse pairs of quotes inside quoted arguments.
                    362:  * Advance the argument pointer to the next argument,
1.71      schwarze  363:  * or to the NUL byte terminating the argument line.
1.36      schwarze  364:  */
                    365: char *
1.48      kristaps  366: mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos)
1.36      schwarze  367: {
                    368:        char     *start, *cp;
                    369:        int       quoted, pairs, white;
                    370:
                    371:        /* Quoting can only start with a new word. */
                    372:        start = *cpp;
1.47      kristaps  373:        quoted = 0;
1.36      schwarze  374:        if ('"' == *start) {
                    375:                quoted = 1;
                    376:                start++;
1.79      schwarze  377:        }
1.36      schwarze  378:
                    379:        pairs = 0;
                    380:        white = 0;
                    381:        for (cp = start; '\0' != *cp; cp++) {
1.67      schwarze  382:
                    383:                /*
                    384:                 * Move the following text left
                    385:                 * after quoted quotes and after "\\" and "\t".
                    386:                 */
1.36      schwarze  387:                if (pairs)
                    388:                        cp[-pairs] = cp[0];
1.67      schwarze  389:
1.36      schwarze  390:                if ('\\' == cp[0]) {
1.67      schwarze  391:                        /*
                    392:                         * In copy mode, translate double to single
                    393:                         * backslashes and backslash-t to literal tabs.
                    394:                         */
                    395:                        switch (cp[1]) {
1.79      schwarze  396:                        case 't':
1.67      schwarze  397:                                cp[0] = '\t';
                    398:                                /* FALLTHROUGH */
1.79      schwarze  399:                        case '\\':
1.36      schwarze  400:                                pairs++;
                    401:                                cp++;
1.67      schwarze  402:                                break;
1.79      schwarze  403:                        case ' ':
1.36      schwarze  404:                                /* Skip escaped blanks. */
1.67      schwarze  405:                                if (0 == quoted)
                    406:                                        cp++;
                    407:                                break;
                    408:                        default:
                    409:                                break;
                    410:                        }
1.36      schwarze  411:                } else if (0 == quoted) {
                    412:                        if (' ' == cp[0]) {
                    413:                                /* Unescaped blanks end unquoted args. */
                    414:                                white = 1;
                    415:                                break;
                    416:                        }
                    417:                } else if ('"' == cp[0]) {
                    418:                        if ('"' == cp[1]) {
                    419:                                /* Quoted quotes collapse. */
                    420:                                pairs++;
                    421:                                cp++;
                    422:                        } else {
                    423:                                /* Unquoted quotes end quoted args. */
                    424:                                quoted = 2;
                    425:                                break;
                    426:                        }
                    427:                }
                    428:        }
                    429:
                    430:        /* Quoted argument without a closing quote. */
1.48      kristaps  431:        if (1 == quoted)
1.42      kristaps  432:                mandoc_msg(MANDOCERR_BADQUOTE, parse, ln, *pos, NULL);
1.36      schwarze  433:
1.71      schwarze  434:        /* NUL-terminate this argument and move to the next one. */
1.36      schwarze  435:        if (pairs)
                    436:                cp[-pairs] = '\0';
                    437:        if ('\0' != *cp) {
                    438:                *cp++ = '\0';
                    439:                while (' ' == *cp)
                    440:                        cp++;
                    441:        }
1.39      kristaps  442:        *pos += (int)(cp - start) + (quoted ? 1 : 0);
1.36      schwarze  443:        *cpp = cp;
                    444:
1.48      kristaps  445:        if ('\0' == *cp && (white || ' ' == cp[-1]))
1.42      kristaps  446:                mandoc_msg(MANDOCERR_EOLNSPACE, parse, ln, *pos, NULL);
1.36      schwarze  447:
                    448:        return(start);
1.4       kristaps  449: }
1.7       kristaps  450:
                    451: static int
                    452: a2time(time_t *t, const char *fmt, const char *p)
                    453: {
                    454:        struct tm        tm;
                    455:        char            *pp;
                    456:
                    457:        memset(&tm, 0, sizeof(struct tm));
                    458:
1.56      kristaps  459:        pp = NULL;
                    460: #ifdef HAVE_STRPTIME
1.7       kristaps  461:        pp = strptime(p, fmt, &tm);
1.56      kristaps  462: #endif
1.7       kristaps  463:        if (NULL != pp && '\0' == *pp) {
                    464:                *t = mktime(&tm);
                    465:                return(1);
                    466:        }
                    467:
                    468:        return(0);
                    469: }
                    470:
1.37      schwarze  471: static char *
                    472: time2a(time_t t)
                    473: {
1.56      kristaps  474:        struct tm       *tm;
1.38      schwarze  475:        char            *buf, *p;
                    476:        size_t           ssz;
1.37      schwarze  477:        int              isz;
                    478:
1.56      kristaps  479:        tm = localtime(&t);
1.37      schwarze  480:
1.38      schwarze  481:        /*
                    482:         * Reserve space:
                    483:         * up to 9 characters for the month (September) + blank
                    484:         * up to 2 characters for the day + comma + blank
                    485:         * 4 characters for the year and a terminating '\0'
                    486:         */
                    487:        p = buf = mandoc_malloc(10 + 4 + 4 + 1);
                    488:
1.56      kristaps  489:        if (0 == (ssz = strftime(p, 10 + 1, "%B ", tm)))
1.38      schwarze  490:                goto fail;
                    491:        p += (int)ssz;
1.37      schwarze  492:
1.56      kristaps  493:        if (-1 == (isz = snprintf(p, 4 + 1, "%d, ", tm->tm_mday)))
1.38      schwarze  494:                goto fail;
1.37      schwarze  495:        p += isz;
                    496:
1.56      kristaps  497:        if (0 == strftime(p, 4 + 1, "%Y", tm))
1.38      schwarze  498:                goto fail;
                    499:        return(buf);
                    500:
                    501: fail:
                    502:        free(buf);
                    503:        return(NULL);
1.37      schwarze  504: }
                    505:
                    506: char *
1.42      kristaps  507: mandoc_normdate(struct mparse *parse, char *in, int ln, int pos)
1.7       kristaps  508: {
1.37      schwarze  509:        char            *out;
1.7       kristaps  510:        time_t           t;
                    511:
1.37      schwarze  512:        if (NULL == in || '\0' == *in ||
                    513:            0 == strcmp(in, "$" "Mdocdate$")) {
1.80      schwarze  514:                mandoc_msg(MANDOCERR_DATE_MISSING, parse, ln, pos, NULL);
1.37      schwarze  515:                time(&t);
                    516:        }
1.62      schwarze  517:        else if (a2time(&t, "%Y-%m-%d", in))
                    518:                t = 0;
1.37      schwarze  519:        else if (!a2time(&t, "$" "Mdocdate: %b %d %Y $", in) &&
1.62      schwarze  520:            !a2time(&t, "%b %d, %Y", in)) {
1.81      schwarze  521:                mandoc_msg(MANDOCERR_DATE_BAD, parse, ln, pos, in);
1.37      schwarze  522:                t = 0;
1.7       kristaps  523:        }
1.37      schwarze  524:        out = t ? time2a(t) : NULL;
1.38      schwarze  525:        return(out ? out : mandoc_strdup(in));
1.7       kristaps  526: }
                    527:
1.12      kristaps  528: int
1.75      schwarze  529: mandoc_eos(const char *p, size_t sz)
1.12      kristaps  530: {
1.75      schwarze  531:        const char      *q;
                    532:        int              enclosed, found;
1.12      kristaps  533:
1.13      kristaps  534:        if (0 == sz)
                    535:                return(0);
1.12      kristaps  536:
1.14      kristaps  537:        /*
                    538:         * End-of-sentence recognition must include situations where
                    539:         * some symbols, such as `)', allow prior EOS punctuation to
1.49      kristaps  540:         * propagate outward.
1.14      kristaps  541:         */
                    542:
1.75      schwarze  543:        enclosed = found = 0;
1.25      kristaps  544:        for (q = p + (int)sz - 1; q >= p; q--) {
1.23      schwarze  545:                switch (*q) {
1.79      schwarze  546:                case '\"':
1.14      kristaps  547:                        /* FALLTHROUGH */
1.79      schwarze  548:                case '\'':
1.15      kristaps  549:                        /* FALLTHROUGH */
1.79      schwarze  550:                case ']':
1.14      kristaps  551:                        /* FALLTHROUGH */
1.79      schwarze  552:                case ')':
1.23      schwarze  553:                        if (0 == found)
                    554:                                enclosed = 1;
1.14      kristaps  555:                        break;
1.79      schwarze  556:                case '.':
1.14      kristaps  557:                        /* FALLTHROUGH */
1.79      schwarze  558:                case '!':
1.14      kristaps  559:                        /* FALLTHROUGH */
1.79      schwarze  560:                case '?':
1.23      schwarze  561:                        found = 1;
                    562:                        break;
1.14      kristaps  563:                default:
1.27      joerg     564:                        return(found && (!enclosed || isalnum((unsigned char)*q)));
1.14      kristaps  565:                }
1.12      kristaps  566:        }
                    567:
1.23      schwarze  568:        return(found && !enclosed);
1.44      kristaps  569: }
1.50      kristaps  570:
                    571: /*
                    572:  * Convert a string to a long that may not be <0.
                    573:  * If the string is invalid, or is less than 0, return -1.
                    574:  */
                    575: int
1.54      kristaps  576: mandoc_strntoi(const char *p, size_t sz, int base)
1.50      kristaps  577: {
                    578:        char             buf[32];
                    579:        char            *ep;
                    580:        long             v;
                    581:
                    582:        if (sz > 31)
                    583:                return(-1);
                    584:
                    585:        memcpy(buf, p, sz);
1.51      kristaps  586:        buf[(int)sz] = '\0';
1.50      kristaps  587:
                    588:        errno = 0;
                    589:        v = strtol(buf, &ep, base);
                    590:
                    591:        if (buf[0] == '\0' || *ep != '\0')
                    592:                return(-1);
                    593:
1.54      kristaps  594:        if (v > INT_MAX)
                    595:                v = INT_MAX;
                    596:        if (v < INT_MIN)
                    597:                v = INT_MIN;
1.50      kristaps  598:
                    599:        return((int)v);
                    600: }

CVSweb